Real world diffraction tests, 40D.

Shot on a tripod, MLU and 2 second timer. Lighting is a single White Lightning strobe. These are very small 100% center crops from full size files. Shot RAW and ran thru lightroom with a slight contrast, about 10 clarity and a very mild sharpen. Lens is a 50mm f1.8 mkI

What do the results tell me? Diffraction exists and is slightly noticeable at f16, and more of an issue at f22.... but still very usable.

Looks sharpest at f5.6. Thanks for the test.

Interesting test epatt.

I'd say looking at these photos I should be more concerned about softness at wide open than diffraction at F22. Admittedly I don't really know what it is I'm trying to see when judging diffraction anyway, but to me 2.8 seems a lot softer than F22 here.
What exactly am I supposed to be looking for when checking for diffraction in any of my photos??

Stop pixel peeping

Thanks for the results.

F/5.6 is excellent, but your test shows that diffraction is not to be feared as many claim.

I never worry about diffraction - I mostly shoot landscapes and small apertures are a must - while there is no doubt that f16 does not deliver the sharpest picture, the results are more than satifactory and certainly look better than having just part of your subject in focus.

Good and interesting test. I'd be really interested to see this test done with a 50D/500D to see how much diffraction is an issue with the higher MP. It should be a bit more noticeble but i'm curious how much.

Very interesting. Clearly diffraction is less of a problem than lens softness wide open.

The slight sharpening during conversion has "reversed/hidden" some of the softening that occurred. The softening happens to all pixels and the sharpening too - you are not seeing all the diffraction "loss" but have also seen how it can be reversed up to a point. All you need to do now is print that composite image at a range of sizes and view the prints as you would normally. You'll see that the range of apertures that look sharp in the prints may be broader than at 100% pixel peeping. You will then be in a position to make your own, real-life tested judgement on when to "worry" about diffraction. So, for example, if you are taking a shot that you know won't be printed large, you can stop way down to 22 (or even further, maybe) but if you know you might want to print a large print of a landscape/sunset/Schna​pps bottle/whatever, you'll know how far you can stop down before seeing an undesirable level of the effect. Who needs theory, when you have your own data for your own workflow?!

artyman wrote in post #8265928
Stop pixel peeping

Certainly not true, if you compare the F2.8 and the F4 you see such a great difference that its worth to shoot it at F4, F5.6 is sharpest imho.

Thanks for the test .

I am still having a hard time with the title of the thread - "Real world diffraction test"

How often does anybody shoot extreme closeups of bottle labels for "real world" shooting? I think if you want to call it real world, use an example of something people shoot where they have to choose between diffraction and DOF.

tkbslc wrote in post #8268099
I am still having a hard time with the title of the thread - "Real world diffraction test"

How often does anybody shoot extreme closeups of bottle labels for "real world" shooting? I think if you want to call it real world, use an example of something people shoot where they have to choose between diffraction and DOF.

I made a test on my neighbor house light about 30 feet away (daytime), at f32 what a mess.

tkbslc wrote in post #8268099
I am still having a hard time with the title of the thread - "Real world diffraction test"

How often does anybody shoot extreme closeups of bottle labels for "real world" shooting? I think if you want to call it real world, use an example of something people shoot where they have to choose between diffraction and DOF.

The point was that this is in contrast to a thread title about a theoretical exercise that could go on for, well, let's say, 35 pages At least I can comfort myself that all that schnapps I drank when I was a lad wasn't from the real world!

Really, though, I agree; he should at least have shot an inclined ruler - I know I like to have the 23 cm mark be sharp in all my photos.

But seriously, though folks, tkbslc has a point in that the apertures when very closely focused may not be the same absolute size as those when focused at infinity (or the distant mountaintop) because the FL changes during focusing. Simple solution. Just move the bottle 20 feet away

tkbslc wrote in post #8268099
I am still having a hard time with the title of the thread - "Real world diffraction test"

How often does anybody shoot extreme closeups of bottle labels for "real world" shooting? I think if you want to call it real world, use an example of something people shoot where they have to choose between diffraction and DOF.

There are quite a few real world applications. Any macro shooter or one taking shots of garden flowers does shoot closeup. However, diffraction has nothing to do with the distance of the subject. Whether shooting close or far, a pixel is a pixel.

gjl711 wrote in post #8268935
There are quite a few real world applications. Any macro shooter or one taking shots of garden flowers does shoot closeup. However, diffraction has nothing to do with the distance of the subject. Whether shooting close or far, a pixel is a pixel.

Exact!